Computing Applications Economic and business dimensions

Life of IP

Seeking to balance intellectual property protection with incentives for investment in innovation.
  1. Introduction
  2. What Is to Be Done?
  3. Proposal 1—Indefinitely Renewable Time Periods
  4. Proposal 2—Exponential Renewal Fees
  5. The Point
  6. Compare to What We Have
  7. Benefits
  8. References
  9. Authors
  10. Footnotes
  11. Figures
Life of IP, illustrative photo

“Intellectual property,” Declared Bill Gates, Microsoft founder and the world’s richest man, “has the shelf life of a banana.” Perhaps. But what if innovation in molecular biology, preservatives, or packaging could give bananas the unrivaled durability of Hostess Twinkies? Those intellectual property life extenders would not only preserve value, they would render Gates’s aphorism a stale anachronism.

The irony is that while any given intellectual property might be as perishable as a banana, IP protections are more like Twinkies. They last longer than might be healthy. The duration of such protections can—and do—privilege creators often at the cost of promoting progress in science and useful arts that the U.S. Constitutiona originally promised.1

IP can be a platform for accelerating economic growth. If skills and know-how are complementary, then adding research to a knowledge base extends new resources to all in the network.4 An exponential increase in patent applications, for example, indicates the growing interest in seeking competitive advantage. Likewise, the rise of code repositories on GitHub, founded in 2008, shows the opportunity provided by code sharing and recombination (see the accompanying figure). Global networks, digital, systems, and interoperable architectures that accumulate innovations just intensify policy debate about the rational balance between appropriate protection and fair use.

As the economics of innovation go exponential, IP protections can prevent more innovation than they protect. Stronger protection increases incentives to invent now but increases costs to invent later.3 Exponential growth then means marginal gains today have magnified costs tomorrow. When policy is more about protectionism than protection, it is inefficient and unwise. We need IP 2.0 policies that respect innovation realities.

Unfortunately, current legislation offers little coherent rationale for how long IP should be protected, whether for mobile apps, life-saving pharmaceuticals, novel materials, or Twinkified bananas. The lengths of IP protections appear divorced from basic economic principles of efficiency and effectiveness. Does policy respect how technological innovation really occurs? Not when there is too much regulatory fiat instead of market mediation. We learned this lesson when we stopped selling airwaves at fixed prices and made them more responsive to markets through spectrum auctions that promoted leaps in innovation. Such market mechanisms can make life—and innovation—better for both buyers and sellers. Innovation increasingly emerges both from “recombination” and “network effects.” Rigid, static, and unresponsive IP regimes that grant maximum protections for fixed lengths of time hurt innovators and customers alike. To paraphrase Clemenceau’s comment about war and generals, IP protections are too important to just leave to the lawyers. A networked world needs markets in IP protection.

Back to Top

What Is to Be Done?

Economists typically remark that optimal IP policies balance complementary tensions: protections must be adequate to stimulate investment without conferring too much monopoly power.5 Balanced protection ensures future competition but not so much that an innovator can’t recover costs and still profit from their innovation investments. But what about opportunity costs associated with potential follow-on innovations in this era of APIs and greater interoperability? Balanced IP policies would recognize and reward successful inventions that result from creative recombination of prior art. When ideas are combined with other ideas, opportunity costs can grow exponentially.7 For example, carbon nanotubes can be used not just for computing devices but also stronger-than-steel sporting goods, flame retardants, body armor, low-friction paints, electronic transistors, heat sinks, solar cells, hydrogen storage, desalination, and delivery of cancer drugs. Innovation is cumulative; greater protection in one area can diminish and/or retard advancement in other areas.1

What are the innovation ecosystem implications? Sooner or later, diminishing returns to individual innovation are overcome by the prospect of increasing returns to collective recombination. A smart and serious IP policy accounts for both market incentives and network effects. IP policy must balance protection with promotion, as the U.S. Constitution declared.

Back to Top

Proposal 1—Indefinitely Renewable Time Periods

Disney has a point in protecting Mickey Mouse—whose shelf life may well exceed even irradiated Twinkies. Certain rare works are so valuable that property rights motivate continued investment and careful husbandry. Landes and Posner, scholars of law and economics, note trademarks can be infinitely renewed to promote high quality and prevent brand confusion.2 If IP terms are infinitely renewable most private works would move to the public domain as property rights owners stopped paying maintenance fees, while a few such as Mickey Mouse would receive continued investment and quality control from extended incentives. IP policies should not be confiscatory but neither should they represent protracted subsidies and free rides.

One-size-fits-all durations are too much like rent control for intellectual property just like those for physical property: government edicts, not markets, set terms. Such price fixing typically limits standards of living and curbs valuable investment.b Dynamic IP 2.0 policies would be fairer, more efficient, and more effective than “rent control.” Inventors could place larger bets on innovation, knowing extension is possible if innovation succeeds. A renewable term would adjust easily across different types of markets such as those for designs, apps, and drugs. Renewal adjusts easily too within a market to designs that endure or fall out of fashion, to books that sell or get remaindered, and to drugs that do or die.

Back to Top

Proposal 2—Exponential Renewal Fees

But infinity is a long time. If longer IP terms increase inventive effort but aggravate monopoly distortion, then infinity implies lots of work for lots and lots of distortion. Appropriate renewal fees must reflect opportunity costs. A testable, rational mechanism could help based on the simple exponential power of combinatorial innovation and network effects.

Suppose the fee starts at just $1, then doubles every year. In year 20, the term under current patent law, the feec would be $542,288. So at a 5% discount rate, the total net present cost would be $436,796.d Current USPTO practice applies renewal fees in years 3.5, 7.5, and 11.5, doubling only twice, and reaching a maximum of only $7,400 for corporations.e

We need IP 2.0 policies that respect innovation realities.

The doubling period could also adjust by property type. For example, copyright fees might apply every third year. Under the 1831 law that allowed for a 42-year maximum copyright term, the final fee would have been $8,192. This is quite reasonable after one has already profited from the work for more than two generations. In contrast, the far-less-reasonable Copyright Term Extension Act (CTEA) grants a term of “life plus 70.” Assuming an author lives for 35 years after producing a creative work, CTEA gives a term of 105 years. Doubling the fee every three years leads to a renewal cost in year 105 of $17,179,869,184. Disney can give the geriatric Mickey a very long life but it becomes increasingly costly to imprison that life in a tower of copyright. Exponentiation does its magic. In contrast to current practice, open renewal allows protection to adjust to true innovation value while exponential fees allow progress to adjust to true innovation costs.

Back to Top

The Point

Duration grows linearly while fees grow exponentially. Benefits and costs must cross. IP will enter the public domain at a time matched to economic value. And, markets—not regulators—decide.

Back to Top

Compare to What We Have

Patent terms date to colonial times. They represent a bargain that might not be rational today. Judge Giles Rich recounts that patent duration started as the length of two apprenticeships.6 Learning a trade as a blacksmith, cooper (barrel maker), or printer required seven years of training. If the apprentice invented anything of value, the master had the right to keep it. This was the master’s bargain for having trained the apprentice; it also kept the master from being put out of business when the apprentice left. Congress extended patent duration to three apprenticeships in 1836. It later compromised at 17 years, between two and three terms, where the law stood for more than a century. U.S. law changed again in 1995, fixing the term at 20 years, to reconcile foreign and domestic practices. This change was not major, given a three-year average review cycle,f because the term timing also changed to 20 years from date of filing from 17 years from date of issue.

Copyright law is no better. Current law provides a term of 70 years plus the life of the author as set by the 1998 Copyright Term Extension Act (CTEA). This law is waggishly known as the “Mickey Mouse Protection Act” based on the extensive lobbying efforts of the Walt Disney Company, which sought legislative action as copyright protection was due to expire. Sonny Bono championed Disney’s cause in the U.S. House of Representatives as his own song “I Got You Babe” also received extended shelf life.

IP policy must balance protection with promotion, as the U.S. Constitution declared.

Many economists, including five Nobel laureates, opposed CTEA on the grounds it could not improve creation incentives retroactively on works already written. Further, CTEA could barely improve incentives prospectively for creators who would have long been dead. These economists filed an amicus brief accompanying the Supreme Court challenge to CTEA.g No friend-of-the-court economists defended it. Yet the Supreme Court ruled 7 to 2 in favor of CTEA, addressing only the limited constitutional question of whether Congress had authority to extend the law retroactively and not whether it was foolish to do so.h Politics won over prudence.

Back to Top


The wisest social policy accounts for reasonable differences in IP duration for boot designs, smartphone apps, and lifesaving drugs. What is reasonable can vary dramatically across functions and across markets. Yet current policy is the same regardless of whether the boots are fads or timeless fashion, or the drugs are snake oil or streptomycin. The option to renew allows investments to adjust to market value. More options can create more innovations.

At the same time, proper policy must balance the promise of innovation against the problem of social costs. Opportunity costs rise increasingly as networks become more connected and ideas can be recombined. Spillovers matter more in a world of bits than one of atoms. Exponentially increasing renewal fees move to match the patterns we observe in an economy of information. The proposed fee structure helps small inventors especially. Fees start very low and increase only as their ideas have time to prove their worth. Annual renewals also address the problem of orphan works—ones protected by copyright but whose authors cannot be found.

As a side benefit, such fees might also attack subterranean and non-practicing patents. Preventing others from using an idea if the owner does not use it becomes increasingly expensive. Geometric fees might not kill a patent troll, but the space under his bridge gets smaller and smaller with each passing year.

In any case, neither slaying patent trolls nor extracting greater fees from IP oligarchs is the point of our reform. The goal is to encourage inventors and IP holders to think more innovatively about how to profit from recombinant innovation and network effects. More efficient and more effective innovation creates opportunities for innovators and customers alike. Let IP protection be as dynamic as the market it serves.

Back to Top

Back to Top

Back to Top

Back to Top


UF1 Figure. Patent applications* 1880–2013 and new software repositories** on GitHub show exponential growth.

Back to top

    1. Heller, M.A. and Eisenberg, R.S. Can patents deter innovation? The anticommons in biomedical research. Science 280, 5364 (1998), 698–701.

    2. Landes, W.M. and Posner, R.A. Indefinitely renewable copyright. The University of Chicago Law Review (2003), 471–518.

    3. Landes, W.M. and Posner, R.A. The Economic Structure of Intellectual Property Law. Harvard University Press, 2009.

    4. Moenius, J. and Trindade, V. Networks, standards and intellectual property rights. Chapter 5 in Frontiers of Economics and Globalization, H. Beladi and E. Choi, Eds., (2007), 157–198.

    5. Scherer, F.M. Nordhaus' theory of optimal patent life: A geometric reinterpretation. The American Economic Review (1972), 422–427.

    6. Schrage, M. Archaic patent laws need to be rewritten. Los Angeles Times (Oct. 24, 1991).

    7. Weitzman, M.L. Recombinant growth. Quarterly Journal of Economics (1998), 331–360.

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More